Sean Hennessy

Key Points: The climate‐medication relationship is bidirectional: climate‐amplified exposures such as extreme heat affect medication access, stability, and safety/effectiveness (downstream), while the use and disposal of medications add to healthcare's carbon footprint, thereby contributing to the very warming that intensifies those exposures (upstream). Downstream: Extreme heat, extreme weather, and supply chain disruption are concrete threats to drug stability and access, from mail‐order degradation of insulin and epinephrine to the Hurricane Helene IV‐fluid shortage, and environmental exposures may alter the effects of common medications. Upstream: Therapeutically equivalent options can differ substantially in carbon intensity; eliminating unnecessary prescribing and selecting lower‐carbon therapeutic equivalents are the levers most directly available to prescribers and pharmacists. The evidence base is still immature: current drug‐heat warnings rest largely on mechanistic evidence and case reports, and advancing the field requires linking environmental exposure data to medication outcomes through case‐crossover and time‐series designs, where confounding by indication and polypharmacy represent central challenges. Where lower‐carbon options are already guideline‐concordant, clinicians and pharmacists can act now through prescribing choices, patient education on heat precautions and proper medication storage and disposal, along with healthcare system preparedness. Broader change requires guideline revision, formulary reform, transparent industy data, and collective action through professional societies, including the International Society for Pharmacoepidemiology.

Estimating treatment effects across disease risk scores (DRS) is a common approach for assessing treatment effect heterogeneity in randomized trials. When external models are unavailable, the optimal approach for internally fitting the DRS model remains uncertain. This simulation study compares three internal derivation approaches, evaluating bias of estimated treatment effects within DRS-defined strata. We simulated trials under varying treatment effects (odds ratios [OR] of 1, 0.8, 0.5) and treatment–covariate interactions. We fit DRS models on 1) controls-only, 2) the full sample ignoring treatment assignment, and 3) a random split-sample of 50% of the controls, who were removed from the second stage of analysis. Additional simulations varied outcome incidence, sample size, randomization ratio, and the true DRS c-statistic. Bias decreased with the split-sample method (overall percent bias (OPB) 7.7% for OR = 0.8 with interactions) compared to the controls-only (OPB = 15.6%) and full-sample methods (OPB = 22.1%). Bias decreased more with split-sample than controls-only and full-sample methods with larger sample sizes, higher outcome incidence, greater treated-to-control ratios, and larger c-statistics. These findings suggest split-sample methods may be the preferred approach to estimate treatment effect heterogeneity by the DRS in trials with sufficient data to support stable prediction modeling.

BACKGROUND: Patients who experience an ischemic stroke while on aspirin therapy present a clinical dilemma about optimal long-term secondary prevention. While switching to an alternative antithrombotic agent is often considered, the effectiveness of switching remains uncertain. METHODS: We conducted a systematic review and network meta-analysis of randomized controlled trials reporting outcomes among patients with ischemic stroke while on aspirin who were either continued on aspirin or switched to an alternative antithrombotic therapy. Alternative antithrombotics included 2 trials of vitamin K antagonists (n=478), 3 trials of dual antiplatelet therapy (n=2229), 3 trials of direct oral anticoagulant (n=2660) monotherapy, and 1 trial of low-dose direct oral anticoagulant added onto aspirin (n=92). We excluded trials of patients with only short-term outcomes of 90 days or fewer, or those with cardioembolic sources of stroke requiring anticoagulation. Our primary outcome was recurrent ischemic stroke; the secondary outcome was a composite of ischemic stroke, myocardial infarction, and vascular death (or all-cause mortality). Outcomes reflect recurrent events measured over a median of ≈19 months (range 11–42 months). In the network portion of this meta-analysis, surface under the cumulative ranking curve rankings and pairwise meta-analyses were used to evaluate and compare the relative efficacy of alternative antithrombotic medications. RESULTS: Data were available from 9 studies (total N=5459 patients) for the outcome of recurrent ischemic stroke. Switching to another therapy was associated with a pooled relative risk of recurrent stroke of 0.88 (95% CI, 0.76–1.03) compared with continuing aspirin, with minimal heterogeneity ( P =0.93; I ²=0). For the composite secondary outcome, 6 studies contributed data, yielding a pooled relative risk of 0.89 (95% CI, 0.72–1.10). In the network meta-analysis, dabigatran, apixaban, and aspirin+low-dose rivaroxaban ranked the highest among antithrombotic alternatives to aspirin, though none were significantly better than continuing aspirin. Rankings were similar when based on posterior estimates from the clinical trials and when using predictive distributions that incorporate between-study variance (ie, expected performance in future settings). CONCLUSIONS: Among patients experiencing ischemic stroke while taking aspirin, switching to an alternative antithrombotic therapy was not conclusively associated with a reduction in recurrent stroke and composite cardiovascular events. Trials are needed to determine whether specific antithrombotic strategies meaningfully improve outcomes in this high-risk population.

Importance: Overdoses involving methamphetamines and cocaine have increased in recent years. Identification of individuals at highest risk could facilitate the implementation of evidence-based interventions to reduce overdose risk.Objective: To develop and internally validate a model that predicts hospitalization or emergency department (ED) treatment for stimulant-involved overdose among the Medicaid-insured population.Design, Setting, and Participants: This was a retrospective case-cohort study using Medicaid claims data from 2016 to 2019 (development) and 2020 (validation) for all Medicaid enrollees age 15 years or older with a cocaine- or other stimulant−involved overdose. A subcohort was created using a simple random sample of the full cohort of all cases. Within the full cohort, cases were identified as those having any inpatient or ED encounter for stimulant-involved overdose during the following year. A case-cohort sample was obtained for each calendar year from 2016 to 2020, each with a subcohort size of 100 000. Each individual contributed only 1 case event (for an individual with multiple overdoses, only the first eligible was selected). For each of the 4 overdose outcomes, a predictive weighted Cox model was first developed among enrollees of sampling years 2016 to 2019 (development set), and its performance was evaluated in our test set of 2020. The prediction models were first developed in November 2023, and the model fairness assessment was performed in April to May 2025.Interventions or Exposures: Individual-level candidate predictors were demographic characteristics, enrollment, health care utilization, and other clinical variables. Area-level variables included social, economic, housing, and demographic characteristics data from the American Community Survey, rural-urban classification, Social Deprivation Index, retail opioid dispensing rates, and health resources.Main Outcomes and Measures: Four types of stimulant-involved overdose associated with hospitalization or ED treatment: cocaine-involved overdose, (1) involving an opioid or (2) not involving an opioid; or methamphetamine-, ecstasy-, or other psychostimulant-involved overdose (hereafter, other stimulant), (3) involving an opioid or (4) not involving an opioid.Results: The analysis included 78 795 enrollees with cocaine- and other stimulant−involved overdose (mean [SD] age, 42.2 [13.7] years; 33 304 [42%] female and 45 491 [58%] male individuals). Weighted Cox regression prediction models showed good calibration and high discriminatory performance (Harrell C statistic): cocaine-involved overdose, with (0.923) or without (0.902) an opioid; other stimulant−involved overdose, with (0.909) or without (0.868) an opioid. For cocaine-involved overdose with opioids, previous individual opioid use disorder diagnosis or cocaine use disorder diagnosis played the largest role in overdose risk prediction. For cocaine-involved overdose without opioids, previous cocaine use disorder diagnosis and area-level income inequality and housing variables contributed most to prediction. For other stimulant−involved overdose with opioids, previous opioid use disorder diagnosis and area-level percentage of those living with a disability contributed most to prediction. For other stimulant−involved overdoses without opioids, previous stimulant-related disorder and area-level proportion of individuals receiving Supplemental Nutrition Assistance Program contributed most to prediction.Conclusions and Relevance: This case-cohort study found that readily available data can be used to identify those at high risk of hospitalization or ED visit for cocaine- or stimulant-involved overdose. These individuals would likely benefit most from evidence-based interventions and awareness of risk factors for overdose.

The cardiovascular risks of exogenous testosterone have been a subject of controversy. In this study, our objective was to examine the association between testosterone (versus glaucoma treatments as an active comparator, with no assumed effect) and the new onset of cardiovascular, cerebrovascular and thromboembolic adverse events, in a US commercial insurance database. Data were analyzed by three complementary designs: inverse propensity score weighting (IPSW), calendar time instrumental variable and instrumented difference-in-differences. Results of these analyses suggest that there is no difference between testosterone and glaucoma treatments regarding the risk of the composite primary endpoint of acute myocardial infarction, ischemic stroke and sudden cardiac arrest/ventricular arrhythmia. In contrast, IPSW analysis identified a negative association between testosterone and the secondary endpoint of venous thromboembolism. However, this association was attenuated towards the null in the calendar time instrumental variable and difference-in-differences analysis, which suggests that there might be unmeasured confounding in the IPSW analysis. Because there is no uniquely suitable method that offers a universally optimal solution for evaluating causal relationships between exposures and outcomes from observational data, using multiple state-of-the-art methods to answer the question of interest can help in assessing the robustness of findings to various forms of unmeasured confounding, thereby aiding in causal inference.

Data from sources other than traditional randomized clinical trials are known as real‐world data (RWD), and the evidence derived from the review and analysis of RWD is known as real‐world evidence (RWE). RWD and RWE are used increasingly throughout the lifecycle of medicinal products to provide evidence about their effectiveness and safety. Recent regulatory guidance about RWE and approvals based on the use of RWE to demonstrate beneficial effects of products have created an urgency to develop generally accepted processes that promote trust in the evidence‐generation process. A recent report from a working group of the Council for International Organizations of Medical Science (CIOMS) describes the use of RWE for decision making in the lifecycle of medical products, describes RWD and data sources, discusses key scientific considerations in the generation of RWE, and discusses ethical and governance issues related to the use of RWD. This paper provides a high‐level summary of this report. More work remains to be done to globally harmonize practices and guidance for using RWD and RWE for regulatory decision making, thereby maximizing the benefits they can bring to patient care and public health.

Characterizing heterogeneity of treatment effects (HTE) is a fundamental goal of pharmacoepidemiology, addressing why medications work differently across patient populations. This paper reviews state‐of‐the‐art methods for studying HTE using real‐world data (RWD), which offer larger study sizes and more diverse patient populations compared to randomized clinical trials. The paper first defines HTE and discusses its measurement. It then examines three leading approaches to studying HTE: subgroup analysis, disease risk score (DRS) methods, and effect modeling methods. Subgroup analyses offer simplicity, transparency, and provide insights into drug mechanisms. However, they face difficulties in resolving which subgroup or combination of characteristics should be the basis for clinical decision making when multiple effect modifiers are present. DRS methods address some of these limitations by incorporating multiple patient characteristics into a summary score of outcome risk but may obscure insights into mechanisms. Effect modeling methods directly predict individual treatment effects, offering potential for precise HTE characterization, but are prone to model misspecification and may not provide mechanistic insights. The methods each have tradeoffs. Subgroup analysis is straightforward but can lead to spurious associations and does not account for multiple characteristics at once. DRS methods are relatively simple to implement and clinically useful, but may not completely describe HTE or provide mechanistic insight. Effect modeling approaches have great potential for characterizing HTE but are still being developed. Understanding HTE is essential for personalizing treatment strategies to improve patient outcomes. Researchers must weigh the strengths and limitations of each approach when using RWD to study HTE.

Insulin secretagogues and angiotensin‐converting enzyme inhibitors (ACEIs) are commonly co‐prescribed for patients with type 2 diabetes (T2D). Case reports suggesting that co‐administration of insulin secretagogues with ACEIs is associated with an increased risk of serious hypoglycemia have led to warnings regarding a drug–drug interaction in widely used drug compendia. However, subsequent studies have had inconsistent results. We performed a systematic review to evaluate the evidence that concomitant use of ACEIs and insulin secretagogues increases the risk of serious hypoglycemia. MEDLINE/PubMed and Embase were searched from inception to July 2023 for studies evaluating adults with T2D treated with insulin secretagogues, such as sulfonylureas or meglitinides, and exposed to an ACEI. The primary outcome was serious hypoglycemia. A literature search yielded 472 papers, of which five met the inclusion criteria. The heterogeneity of the studies precluded meta‐analysis. Two studies using multiple methods to address bias found no association between hypoglycemia and concomitant use of ACEI and insulin secretagogues. Three studies found potential associations, but only one was statistically significant; these studies were at serious or critical risk of bias due to potential confounding from lack of adjustment for renal dysfunction. The higher quality studies found no association between the concomitant use of insulin secretagogues with ACEI and hypoglycemia. Drug compendia and electronic health records should consider updating and removing alerts warning of a drug–drug interaction between insulin secretagogues as a class and ACEIs.